Latest / Connor with Honor: AI, Real Estate and Real Life / AI Got the Math Right and the Table Wrong. Test Before You Trust It.

Transcript

AI Got the Math Right and the Table Wrong. Test Before You Trust It. · Connor with Honor: AI, Real Estate and Real Life

• 0:01
  SPEAKER_00: The house next door to you might be a terrible comparison for yours.
• 0:06
  SPEAKER_00: Same street doesn't mean the same buyer would choose it.
• 0:09
  SPEAKER_00: Now imagine your home has stairs and a small yard.
• 0:12
  SPEAKER_00: Another home nearby has everything on the ground floor and room for a pool.
• 0:17
  SPEAKER_00: Those differences actually may matter more to a particular buyer than the distance between the front doors.
• 0:23
  SPEAKER_00: Before I call a listing on your competition, I want to understand the choice somebody is making beforehand.
• 0:31
  SPEAKER_00: Can the home solve similar needs in a similar budget?
• 0:34
  SPEAKER_00: How do the layout, condition, outdoor space, and recurring expenses compare?
• 0:39
  SPEAKER_00: A beautiful kitchen does not erase a floor plan that fails the buyer's needs.
• 0:44
  SPEAKER_00: An attractive asking price does not tell us the monthly ownership costs include.
• 0:49
  SPEAKER_00: I'd build a short comparison list with a reason besides each property.
• 0:53
  SPEAKER_00: Why would a buyer consider this instead of yours?
• 0:55
  SPEAKER_00: What advantage does it offer or have?
• 0:58
  SPEAKER_00: What does your home offer?
• 0:59
  SPEAKER_00: Does it offer something different?
• 1:01
  SPEAKER_00: And then we can decide what your photographs and description need to be made clear.
• 1:06
  SPEAKER_00: Now that's more useful than collecting nearby addresses and calling the job finished.
• 1:11
  SPEAKER_00: I represent sellers and understanding the buyer's alternatives as part of what work.
• 1:16
  SPEAKER_00: Now, if you're preparing to sell in Santa Clarita, visit sellersonlyagent.com to start a conversation about your property.
• 1:28
  SPEAKER_00: I gave AI a simple property comparison and it got the math right while getting part of the table wrong.
• 1:35
  SPEAKER_00: And here's that test using fictional homes.
• 1:38
  SPEAKER_00: Property A has an association due of$1,200 per year.
• 1:42
  SPEAKER_00: Property B has a due of dues of$300 per quarter.
• 1:46
  SPEAKER_00: Property C has no dues, information supplied.
• 1:50
  SPEAKER_00: Now the job was to put those amounts on the same monthly basis and identify missing information.
• 1:56
  SPEAKER_00: The correct monthly equivalent is$100 for A and$100 for B for C.
• 2:01
  SPEAKER_00: Well, the answer is unknown.
• 2:03
  SPEAKER_00: The blank field doesn't mean the property has no association dues.
• 2:07
  SPEAKER_00: The AI returned those monthly amounts correctly, but in its missing information column, it also marked the dues for A and B as missing as well.
• 2:15
  SPEAKER_00: And they were right there in the input.
• 2:17
  SPEAKER_00: That error is easy to overlook because the table looks organized.
• 2:21
  SPEAKER_00: Now I checked the numbers with division and checked each missing label against the original notes.
• 2:26
  SPEAKER_00: Special assessments were not supplied for any of these fictional properties, so those need checking as well.
• 2:33
  SPEAKER_00: Now, there's a useful way to organize property information when somebody actually reviews the result.
• 2:38
  SPEAKER_00: The table is not a complete ownership budget.
• 2:41
  SPEAKER_00: It can't tell us where that home is worth more.
• 2:44
  SPEAKER_00: Try this.
• 2:45
  SPEAKER_00: With fictional information, give AI the original amounts and payment periods, then check every row before relying on it.
• 2:55
  SPEAKER_00: Can asking prices fall while sold prices rise?
• 2:59
  SPEAKER_00: Yes.
• 3:00
  SPEAKER_00: And these reports show while the lay why the label besides that number matters.
• 3:04
  SPEAKER_00: Realtor.com September 2026 report puts the national median asking price at$419,250.
• 3:12
  SPEAKER_00: That's down 1.4% from September of 2025.
• 3:17
  SPEAKER_00: The National Association of Realtors reported in August 2026 median existing home sales price of$429,100, up 1.6% from August 2025.
• 3:30
  SPEAKER_00: Those are different measurements covering different months and different groups of homes.
• 3:34
  SPEAKER_00: An asking price describes what a seller is seeking.
• 3:37
  SPEAKER_00: A closed sale record, a transaction is what actually finished.
• 3:42
  SPEAKER_00: Now you can't subtract those national medians and announce what buyers negotiated.
• 3:47
  SPEAKER_00: They're not matched asking and selling prices for that same properties.
• 3:52
  SPEAKER_00: Now the median is the middle of the group, and the mix of homes in that group can change.
• 3:57
  SPEAKER_00: More expensive homes entering the sold sample can change that middle without proving that every home gained value.
• 4:04
  SPEAKER_00: For a Santa Clara homeowner, this is the reason to inspect the evidence before changing the plan.
• 4:09
  SPEAKER_00: I would want recent comparable home sales, the current alternatives, and the details of your property.
• 4:15
  SPEAKER_00: National reports provide context, but they don't appraise your house.
• 4:18
  SPEAKER_00: Before a housing headline changes your next move, check the place, the reporting month, and whether the number describes asking prices or completed sales.
• 4:34
  SPEAKER_00: Your AI agent can sound familiar and still leave your customer stranded.
• 4:38
  SPEAKER_00: Now, before you put it on your business line, you want to test what happens when it can't answer.
• 4:44
  SPEAKER_00: Use a fictional caller in a situation your team actually recognizes.
• 4:48
  SPEAKER_00: Ask about a service that you offer.
• 4:51
  SPEAKER_00: Then ask for something outside the information you approved.
• 4:54
  SPEAKER_00: Finally, ask for a person.
• 4:56
  SPEAKER_00: Now you're checking the next step.
• 4:59
  SPEAKER_00: You're not just whether that voice sounds natural.
• 5:02
  SPEAKER_00: Does it explain the limit?
• 5:04
  SPEAKER_00: Does it capture the request accurately?
• 5:06
  SPEAKER_00: Is somebody responsible for following up?
• 5:08
  SPEAKER_00: A message in the dashboard doesn't mean a person has seen it.
• 5:12
  SPEAKER_00: And collecting a preferred appointment time doesn't mean the appointment is actually booked.
• 5:16
  SPEAKER_00: That calendar connection and confirmation need to be tested before you promise that behavior to your customers.
• 5:23
  SPEAKER_00: Now, for a local service business, I would start with a narrow job such as after hours intake.
• 5:29
  SPEAKER_00: Agree on the questions it could answer and information it should collect, and where a usual request goes.
• 5:36
  SPEAKER_00: Review what happened on actual calls before expanding the job.
• 5:39
  SPEAKER_00: That's the kind of system I want to build with Higher AI Voice.
• 5:43
  SPEAKER_00: Now visit higherAIVoice.com to explore how it works and talk through where your calls get stuck in your business.
• 5:56
  SPEAKER_00: If reading the news leaves you with more outrage than understanding, try changing the format.
• 6:02
  SPEAKER_00: That's why I built the Honor Daily.
• 6:05
  SPEAKER_00: I want to sit down, understand what happened, and finish reading.
• 6:07
  SPEAKER_00: A headline alone rarely gives me enough to do that.
• 6:10
  SPEAKER_00: I want the source, the context, and a clear distinction between what is established and what somebody thinks it means.
• 6:17
  SPEAKER_00: The paper brings together world news, property, artificial intelligence, health, and ideas worth considering.
• 6:24
  SPEAKER_00: You can follow the underlying sources instead of taking a summary on faith.
• 6:28
  SPEAKER_00: For property coverage, pay attention to the place and the date attached to the numbers.
• 6:33
  SPEAKER_00: For an AI story, ask what was demonstrated and what is still a promise.
• 6:38
  SPEAKER_00: For a health claim, look at who was studied and the evidence and what the evidence actually supports.
• 6:43
  SPEAKER_00: Those questions help you decide whether a story deserves your attention.
• 6:47
  SPEAKER_00: Now you can read the paper before deciding whether you want it in your inbox.
• 6:51
  SPEAKER_00: The sign up page lets you choose a reading frequency and explains the current delivery status.
• 6:57
  SPEAKER_00: I want that paper to fit your day, whether you follow it closely or check in less often.
• 7:02
  SPEAKER_00: Go to BookwithHonor.comslash newsletter to read more and choose your pace.